CHAPTER 1: Introduction
Visual inspection is defined as a quality task that determines if a product deviates from a given set of specification using visual data. inspection usually involves measurement of specific part feature. if the measurement lies within a determined tolerances ,the inspection process considers the product as accepted for use.

In industrial environment inspection is performed by human inspectors or visual inspection system. Human inspectors are not always consistent and effective evaluators of products because inspection tasks are monotonous and exhausting .typically there is one rejected in hundreds of accepted products. it has been reported that human visual inspection is at best 80% effective.
In addition achieving human 100% inspection where it is necessary to check every product thoroughly, typically requires high level of redundancy thus increasing the cost and time for inspection. For instance, human visual inspection has been estimated o account for 10% or more of the total labor costs for manufactured products .for these reasons in many application batch inspection is carried out. in this case a representative set of products is selected in order to perform inferential reasoning about the total. Nevertheless in certain application a 100 % inspection is required.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

This project is actual requirement of “HINDUSTAN UNILEVER LIMITED” which is given to eastro control system to find out the solution. They want high accuracy with quality inspection, less man power, FuIIy Automated system, no manual intervention, improve production process with higher speeds, without defects. so instead of doing this work manually we find the solution using vision system.

“HINDUSTAN UNILEVER LIMITED” has wished to measure all parameter related to ICE-CREAM PACK like date of ice cream pack, flavour of ice cream pack and price too. Hence on this project ESTRO CONTROL SYSTEM start working.

TARGET COMMUNITY OF PROJECT: Inspection of components using machine vision technologies provides solution for quality and process control. Various applications of Machine vision technologies are automotive, Pharmaceutical, food and beverage, electronics, packages, process control and special application. In this project optical character reorganization, process control using image processing, checking of presence and absence of finished product parts in the production line are checked. We can used such systems in those industries where inspection is needed (for example in agri industries ,medical industries ,traffic controlling etc.)
The future scope of this automated application is in the industry of harvest engineering and Automated Agriculture.

This automated application will be helpful for easing of supply chain management.

This application is used to detection of bad quality tomatoes using computer vision.

It is also used to analyze remote sensing data for farming purpose and larger scale control production.

Image processing technique has been proved as effective machine vision system for agriculture domain. we can conclude that image processing was not invasive and effective tool that can be applied for the agriculture domain with great accuracy for analysis of agronomic parameters.

Quality Improvement of product.

Sort Out the Defective Products.

Check continuously manufacturing date, cost and flavor of ice cream pack.

Reduces The Need Of Man Power.

Production Process Becomes Faster
Higher Accuracy & Requires Less Time .

Automated system, no manual intervention.

Many key tasks in the manufacture of product, including inspection ,orientation, identification and assembly, require the use of visual technique .nothing beats human vision of versatility , but human weakness limit productivity in a manufacturing environment .boredom, distraction, fatigue, and sometimes even malice can degrade human performance in vision related factory tasks such as inspection. Factory automation utilizing a machine vision system in such tasks can bring many benefits.

Machine vision system can perform can perform repetitive task faster, more accurately, and with greater consistency over time than humans. They can reduce labor costs, increase production yields and eliminate costly errors associated with incomplete or incorrect assembly. they can help automatically identify and correct manufacturing problems on line by forming part of the factory control network. The net result is greater productivity and improved customer satisfaction through the consistent delivery of quality products.

Before implementing the vision system inspection done manually which is time consuming and chances of error in inspection is maximum, Like ice cream pack ,bottle filling ,and so on multiple project done manually .In “HINDUSTAN UNILEVER LIMITED” firstly all works done manually like check the ice cream pack flavour manually ,counting of ice cream ,date of manufacturing , cost of product, etc. In all these work time consuming more and no of employee requirement also more and speed of manufacturing of product is also less. Hence to avoid these problems “HINDUSTAN UNILEVER LIMITED” start working on vision system.

Hence “HINDUSTAN UNILEVER LIMITED” sponsored project to ESTRO CONTROL SYSTEM under the estro control system we done this project.

Many key tasks in the manufacture of product, including inspection , orientation, identification and assembly, require the use of visual technique. Nothing beats human vision of versatility , but human weakness limit productivity in a manufacturing environment .boredom, distraction, fatigue, and sometimes even malice can degrade human performance in vision related factory tasks such as inspection. Factory automation utilizing a machine vision system in such tasks can bring many benefits.

Machine vision system can perform can perform repetitive task faster, more accurately, and with greater consistency over time than humans. They can reduce labor costs, increase production yields ,and eliminate costly errors associated with incomplete or incorrect assembly. they can help automatically identify and correct manufacturing problems on line by forming part of the factory control network. The net result is greater productivity and improved customer satisfaction through the consistent delivery of quality products.

There are a number of critical questions to ask up front:
What tasks does the system need to perform?
What are the keys visual performance criteria?
What are the environment factors?
Who will program the system?
What equipment must the vision system interface with?
What information must the system provide?
What are the operator requirements?
System type:
Embedded / vision engine.

Modular/pc- based
Inspection rate:
Product speed.

Number of cameras.

4. Processor speed.

5. Data storage.

6. Applications/capabilities.

Electronics or semiconductor inspection.

7. Image source.

Area scan camera.

High speed.

8. View suppliers by state.

9. View datasheets by specs.


Triggering sensor – Proxy retro-reflective :This sensor is PNP NO type when the object is sensed it gives the signal to vision system through PLC as a trigger signal to capture an image by camera. This sensor has a sensing distance 3-meter maximum.

PLC – CP1L-EL20DT(IP Address –
Shift logic and vision triggering programming are done in PLC. PLC has its own RTC, we can configure it as we require.

Vision Controller – FZ5-L355 :
It is PNP type controller with two camera port. In this, we can do programming for each flavor. Here we are done programming for flavor detection and proper printing of coding is done or not with respect to flavor. The controller performs anoperation on the imagecaptured by the camera and give OK or NG signal. This signal is going to tower lamp through PLC for indication and to the cylinder for rejection of NG (not good) party pack.

Camera – FZ-S:
Two cameras are used -First is mounting in a vertical position for detection of coding.

Second is mounting in a horizontal position for detection of flavor.

It displays the two images captured by two cameras for visual inspection and for programming on images.

Cylinder and rejection assembly with conveyor
When a controller gives NG signal, the cylinderis getting actuate and reject the not good pack which is dropped into rejection bin At home position of the cylinder, there is a reed switch.

Tower lamp: Here we use three-tier hard wire indication lamp. This indicates OK, NG and error signals which are come from vision controller.

On door buttons and indications:
Vision on/off- It controls the triggering signal if it is on then triggering signal read by vision controller and if it is off the triggering signal not read by vision controller.

Camera rejection 1 – When horizontal camera gives NG result it glows high.
Camera rejection 2 – When vertical camera gives NG result it glows high.

Rejection on/off – It bypasses the NG signal given to the cylinder for rejection.

Drive start – Conveyor start.

Drive stop – conveyor stop.

Drive trip.

This System Is Basically Design For The Automation Of Industry. Here We Are Using Omron Vision Controller It Offers High-Speed Input Response Of 0.1 Ms And Equipped With Built-In Timer In Order To Count No. Of Ice-cream pack. With The Help Of Controller & Camera Here We Are Check Ice cream pack manufacturing date ,It’s price., Whether The Flavor is correct or not. Cameras will Continuously Checking
All These Things ; Send the Signal to Controller to take the Proper Action. Here Monitor Is Used to display The Count as well as to Display The Message When Defective Ice-cream pack Comes. We Are Using the Proxy retro-reflective sensor to Sense the Object (Ice-cream pack). Tower Lamp Is Used For Indication Purpose That Whether The Ice-cream pack Flavor is Matched With Reference Or Not. If it is matched then The Lamp Will Be Indicate the green light. If It Is Not Matched Then It Will Be Indicate red light. If there is a error in wiring connection that time orange light is glow. Conveyer System is used for moving the Ice-cream pack.
PLC used for rejection mechanism If coding is different than respective flavor then vision system will give NG signal to PLC and reject the party pack. If coding is not present, then it will reject the party pack. If the carton is of another flavor than running line flavor, then it will reject. If coding is done on opposite face of party pack, then it will reject. If coding is not done properly because of water droplets or ice-cream on the surface, then it will reject.
Hardware design:
Vision controller-FZ5-L355:
Four core image processing.

Fast ethercat communication.

Up to high resolution 8 camera.

High speed digital CMOS.

Memory at least 2 GB RAM.

Power supply voltage 20 V to 26 VDC.

Cooling operating:0 to 50 degree Celsius fan speed.

Current consumption :0.7A max
Parallel I/O – either an NPN or PNP output is supported.

Ethernet and RS-232 connectivity.

Compatibility with FZ and FH cameras.
Image acquisition time :8.5ms to 2ms.

Model No. FZ-30254M.

High speed digital CMOS cameras (lens required).

4 million pixels.

Colour or monochrome.

Camera cable : FZ-VS3 (2 meter).

Frame Rate : 80fps.

Lens Mounting: c mount.

Conveyor selection:
Load: 1 to 2 KG .

Distance: 1 to 1.5 m long.

Width: 20 cm.

Ac motor: 0.5 HP
Ac drive : variable frequency drive.


Serial communication: two port(either Rs-232C or Rs-485)
High speed counters, single phase for four axes.

Six interrupt inputs are built in.

Faster processing of instruction speed up the entire system.

LCD display and setting. Enabled using option Board.

In Omron plc we install XP operating system.

5. Proxy retro-reflective

Dimension: 55*50*20mm
Range : 3 meter
Housing: ABS plastic
Output: PNP or NPN-4 wires
Output type: NO or NC selectable with sensitivity adjustment.

Supply: 10-30 VDC.

Output current: 200 MA.

Protection: short circuit reverse polarity.


Operation: Steady/flashing.

Power supply: AC/DC 24V
Buzzer sound level: fixed-85dB
Lens color: 3 colors in 1 module-RED/BLUE/GREEN.

Lens diameter: 56 mm
Maximum tier: 2
Degree of protection: NEMA 4x.IP65/IP50.

Vibration: 3mm at 10-55Hz.

MODERN tools used
CX programmer is the programming software for all Omron PLC series. It fully integrated into the CX-one software suites.CX programmer include a wide variety of features to speed up the development of your PLC program. New parameter setting dialogues reduce setup time and with standard function block in IEC 61131-3 structured text or conventional ladder language, CX programmer makes development of PLC programs a simple drag and amp; configuration.

Special data types for TIMER (count-down) and COUNTER (count – up) symbols greatly simplify the use of timer/counters in ladder program as rung; to reset and check them you can simply access them by using their name. When used with the Auto allocation feature, you can define a symbol of type TIMER or COUNTER and never have to worry about where it is stored. That means zero maintenance to resolve addresses when a program grows or rungs are copied to a new project .Arrays of timers and counters are also supported.

This software’s manual describes the basic operations and all settings when using vision sensors. This Manual includes two parts: Operation and Environment setting as follows.

Operation Part:- First, prepare to measure by performing “Preparation” -> “Design Contents to be Measurement” -> “Set up Contents to be Measurement”. Then, confirm whether or not to measure exactly against the expected objectives by performing “Measurement Test” -> “Adjusting Measurement Contents” repeatedly. Finally, perform “Actual Measurement” -> “Verify Measurement Results”. In addition, this part also describes the operations of controller and how to utilize and analyze the measurement results.
Environment Setting Part:-This part describes the measurement environment (camera and lighting, etc.) and the required system setting and configuration to facilitate the measurement. In addition, this part also describes the required setting for outputting the measurement results to external devices and how to store data.

Capture image from camera.

Identify the shape.

Perform position compensation.

Check for defects.

Output measurement results to PLC.


In modular construction up to 70% of the value of project is in manufactured component that are delivered ‘just in time’ to site. The off-site construction process leads to considerable sustainability and site organizational benefits, such as
Significantly smaller carbon footprint than traditional construction methods due to the speed of construction resulting in overall project being delivered at least 40% quicker than traditional construction.

Reduced storage and hire charge.

Construction work on site is inherently dangerous. with vision Modular system, process can be carried out in more easily met and policed, and healthy and comfortable working conditions are more readily maintained.

Reduced waste and landfill charge, and more opportunities for recycling. There is 90% less waste produced on site over traditional build.

Greater recycling of material in factory condition by partnering with specialist waste reduction advisors and employing specialist waste management contractors.

Exploding beer bottles, tsunamis and flour may not seem to have anything in common, but they are all hazards that machine vision systems can face in field applications.

System designs targeting the controlled condition of a lab expect an ideal working environment, not real world problem.

For most imaging applications, temperature challenges revolve around dissipation of heat generated within the equipment.

Field environment rarely are cold enough to interface with the operation of the electronics; however, these easily become hot enough to cause equipment failure when there is inefficient heat dissipation.

Ideally, the electronics, especially the image sensor, should be kept below 50 degree.

Machine vision system use heat sink and fans for controlling heat.

5. Vibration is the leading cause of a number of failure, including alteration of a number of failures, including alteration of the imaging lens focus, unthreading of the lens from camera, optical misalignment ,and electronic equipment joint and connection failures.

6. Shock also present in many installation particularly in machinery that is engaged in repetitive motion and can have similar effects.

7. Water and particulates:
Exposure to water or other fluids can wreak havoc on an imaging system by causing short circuits in the electronics, resulting in equipment failure.

Fluids also can harm lenses by damaging coatings or by creating
Condensation that can impair resolution and light transmission.

Avoiding fluid damages on where the fluid comes from. For
Incident fluids such as rain, merely deflecting the fluid may be adequate protection.

1. Conditions for OK images.

No Image Description
1. 245110142875 Inspection of coding
Coding is present.
Coding is must be with respect to flavor and shift.

Proper coding must be done.

2. -23495462915 Inspection of flavor
If chocolate flavor programme is selected, then pack should be of chocolate flavor.

If it is other than selected flavor, then it will reject.

2. Conditions for NG images.

No Image Description
1. 10160343535 Inspection of coding.

1. Proper coding is not done because an ice-cream is present in the coding area.
2. 67310389255 Inspection of coding.

1. Coding is done on opposite face of the pack so, it will reject.

641985569595Programming for Party Pack Vision System
Function Block- 2 Camera Switching
This shows the image of the horizontal camera which inspecting the flavor.

Function Block- 3 Shape Search II
This block inspects the flavors. For ex.chocolate. In the model region, we can set the interest area for inspection.

Following are the steps which need to follow for setting the model.

d) Measurement parameter.

3.Function Block- 6 Camera Switching
This shows the image of the vertical camera which inspecting the coding.

Function Block- 7 Background suppression
This block used to separate background from the inspected area so we will improve the visualization.

Function Block- 8 Shape Search III
This block is used to give a reference point to the OCR block for inspection of coding. As it is detected RS.125 and this gives reference to OCR block.

Follow the steps as per mention in function block 3 descriptions.

Function Block- 10 DI Branch
This block helps to run OCR functions with respect to shifts. As we are using 3 OCR for 3 shifts here we set the condition to use OCR with respect to shift. For ex. For 1st shift OCR, 11 is selected and so on.
Function Block- 11 OCR
OCR means Optical Character Recognition; here we configure it to inspect coding on party pack with respect to flavor for 1st shift. In the model region, we set the area in which coding is printed.The set area must be larger than printed area. In judgment condition, we enter the character which we want to inspect.

For ex. RS.125/- 10/1/18 A10FS, if read character match to this targeted character then results is OK and if not then the result is NG (Not Good).

Please follow the following steps for setting the OCR.





Select OCV in inspection mode.

In judgment condition enters the characters which we going to inspect in block line No. 1. For ex. RS.125/- 12/03/18 A12FS
Click OK.

Function Block- 12 OCR User Dictionary
Here we register the character 0 to 9 which are we going to inspect. Here we register 10 different patterns for each character.
Function Block- 13 Calculation
In this block, we can set the count for total No. of party packs and NG images for 1st shift.
10. Function Block- 14 Parallel Judgement Output
Results of function block 3 and 11 are assigning to on door indications which are camera 1 rejection and camera 2 rejection respectively.
11. Function Block- 15 Image Conversion Logging
For logging NG (Not Good) images, we use this block. Here we select the only NG image logging option, set the destination folder in which images are going to be store and set area of the image for logging.

12. Function Block- 16 End
This block used to end the branch and after this new branch is started.

Note- From function block 11 to 16 is repeated for 2nd shift and 3rd shift programming. OCR 17 and 23 is for inspection of coding in a 2nd shift and 3rd shift.

CONCLUSION : Vision is best for qualitative interpretation of a complex, unstructured scene, machine vision excels at quantitative measurement of a structured scene because of its speed, accuracy, and repeatability. The basic idea of this project is to make aware of machine vision technology and to improve the production quality, reduce the scrap product due to non-conformity by controlling the manufacturing process through machine vision and also to prevent the value addition for scrap product in the subsequent stage of manufacturing process.

With the help of higher quality vision system this project is implemented for inspection of ice-cream pack . Machine vision can lead to significant cost reduction. Cost of this project is also reduced because of machine vision.

We can increase the speed of production line by increasing the no. of cameras to the controller.

We can connect upto8 cameras to each controller it’s depend on which series of fz controller is used.
At time we can controls multiple no of conveyors.

With the help of faster hardware and more intelligent software this system can be used for many industries.